Horizontal Prism Tolerance Question

[Aarlan]

Hey all,

I have a quick question. It has been my understanding that 0.67 prism diopters is the Horizontal imbalance tolerance (as long as < 2.75 D), whether it is 0.67 in one eye, or any combination in the two lenses together, as long as it is not over 0.67 TOTAL


I was discussing with an instructor, and they claim that the tolerance is 0.33 horizontal prism MAX EACH eye, up to a total of 0.67. I've looked and looked, but I haven't found the 0.33 D max each eye horizontal requirement anywhere, only the 0.67 TOTAL.


Am I missing something?

AA



PS...I agree that 0.67 in one eye is a sign of poor workmanship, and would question the lab that came from, but for argument's sake I want to know if I'm off base here.

[lensgrinder]

You are correct, the tolerance is 0.67 D of prism total for both eyes combined up to a 2.75 D and 2.5mm above 2.75 D.

I think where that might be coming from is in Optical Formulas Tutorial it talks about putting 0.33 D of prism in each eye and re-measure the PD to see if the PD is then correct.

[HarryChiling]

Replace the mounted lens pair in the focimeter, centering the prism reference point of
the lens with the strongest refractive power through the horizontal meridian in front of
the lens stop. If the measured distance from Step 2 (the distance between the
positions at which the prismatic requirements of the prescription are met) is wider than
the specified interpupillary distance, slide the lens out until 1/3

∆ is induced. Mark the
lens at this position. Repeat this for the other lens. If the measured distance is
narrower than the specified interpupillary distance, slide the lens in from the lens stop
instead;...............................................

Figure 4 – Horizontal Prism Tolerance Markings

The actual distance between optical centers (OC) is narrower than the specified
interpupillary distance (PD). Outside markings have been placed where 1/3

∆ has been
induced. Because the outside markings are wider than the PD, the pair passes the
horizontal requirements.

1/3 prism is the tolerence horizontally in each eye, or 0.67 total for both eyes, but I have never heard of 0.67 in one eye and 0.00 in the other as being aceptable and ANSI clearly dictates that it would fail acording to standard.

[lensgrinder]

What ANSI is showing is a method to determine whether the total amount exceeds the 0.67 tolerance without using prentice's rule. The standard reads 0.67 total, not 0.33 each eye.

[HarryChiling]

Lensgrinder,

Your right the method and standard does allow for 0.67 Prism Diopters to be in one eye and zero in the other, I have never allowed that to pass through my lab, and it has never been very difficult to maintain a higher standard. Good catch, but it was intended to be 0.33 in each lens. If you look at the 1964 ANSI standards the standard was clear in that the tolerence was for each lens, I have attached an image from it.

[HarryChiling]

I have never heard anyone bring this up before and it is another case where the standard has been made lax. This was definately a more subtle change as the sum total has only changed by 0.12 prism diopters the smallest increment of measure we use, however if the 0.67 prism diopter applies to a single lens than that woul dmore than double the original standard. This would make for a great memo to the ANSI commitee to revise and make clearer the intentions of the prism tolerence in the next revision.

[Aarlan]

1/3 prism is the tolerence horizontally in each eye, or 0.67 total for both eyes, but I have never heard of 0.67 in one eye and 0.00 in the other as being aceptable and ANSI clearly dictates that it would fail acording to standard.

Harry,

Every chart I have ever seen states the following:

Horizontal Prism Imbalance

IF ≥ 0.00 D, ≤ &#177;2.75 D....then &#177;0.67 Δ Total
IF > &#177;2.75 ...................then &#177;2.5 mm Total



There is no mention of 1/3 for each eye in any ANSI chart of table I have ever seen (exept in the exerpt you posted) it just states total. Now in our office we use 1/3 each eye because we choose to, but to be precise, if you are teaching and/or tutoring for testing purposes it seems that 0.67 total (either eye) would technically be the correct answer, not 1/3 each eye up to 2/3.

Take for example OD -7.00 sph
OS -1.00 sph PD 30/30
if the pds were 30/34, then using prentices rule there is no prism OD and .4 D horizontal OS. This appears to be OK using the 0.67 total method, but not the 0.33 each method. Which is correct?

AA

[Aarlan]

And how about the 2.5 mm total for higher power lenses? Does that mean for a 7.00 sph OU that one lens can be a 2.5 mm off, or each eye can only be off 1.25 up to a total of 2.5 total?

Although I would not want one eye to be off 2.5 mm from where I specified, it seems that a literal reading of the standards allow this as well, assuming the other is absolutely correct.

AA

[HarryChiling]

Aarlan,

Again although the text does not literally spell it out it was intended in the first ANSI standard and the meaning might have been lost in the translation, if you look at the images above, the first is from the current standard and the second is from the 64' version of the standard. I am going to look over all the standards and see if I can find where the text became ambiguos about prism tolerance. I will post my findings.

[Darryl Meister]

You both are sort of right. The current test method, and the original standard, was written with 1/3 prism diopters per eye in mind. However, since either the 1979 or 1987 standard (I forget which), the standard was changed to 2/3 prism diopters or 2.5 millimeters of error, total -- which may or may not be split between the two lenses. However, the test method has remained largely unchanged, for several reasons.

Technically, you are now allowed to have up to 2/3 prism diopters or 2.5 millimeters of error in only one lens or split between both -- at least for single vision and bifocal lenses. The standard was changed back in 1999 to allow only 1.0 millimeter of error per eye for progressive lenses. At the end of the day, keep in mind that it doesn't matter to the wearer's visual system whether the total prism imbalance is in one lens or split between both for this type of error.

[Aarlan]

So if you are teaching for the ABO, or for training on a standardized test, how would you approach it?

I would assume from what you said that for progs it is 1/3 ou up to 2/3 total, but for all others it is 2/3 total.

AA

[Darryl Meister]

So if you are teaching for the ABO, or for training on a standardized test, how would you approach it?

Honestly, I couldn't say whether anyone from the ABO has even revised the test questions since the 1999 and 2005 updates. However, I would teach the tolerances, not the test methods.

I would assume from what you said that for progs it is 1/3 ou up to 2/3 total, but for all others it is 2/3 total.

My original post was a little misleading here (so I've taken the liberty of rewording it slightly). Specifically, the centration part of the tolerance is now 1 millimeter per eye for progressives, instead of a combined 2.5 millimeters -- as it still is for single vision and bifocal lenses. Theoretically, a pair of low-powered progressive lenses could still have 2/3 prism diopters of unwanted horizontal prism in a single lens. Here is a summary of the Z80.1-2005 tolerances.

[msparks7]

[quote=Aarlan;193555]Hey all,

I have a quick question. It has been my understanding that 0.67 prism diopters is the Horizontal imbalance tolerance (as long as < 2.75 D), whether it is 0.67 in one eye, or any combination in the two lenses together, as long as it is not over 0.67 TOTAL


I was discussing with an instructor, and they claim that the tolerance is 0.33 horizontal prism MAX EACH eye, up to a total of 0.67. I've looked and looked, but I haven't found the 0.33 D max each eye horizontal requirement anywhere, only the 0.67 TOTAL.


Am I missing something?


Check out this line in ANSI Z80.1-2005
Unmounted Prism and PRP ≥ 0.00 D, ≤ ±3.37 D
> ±3.37 D 0.33 Δ
1.0 mmI take it they mean 0.33 tolerance per eye horizontal and vertical


:cheers:
I'll drink to anything

[bt5050]

Hi everyone -

Not to sound stupid - but wanted some clearification of Prism in a PAL -

When they are referring to 1mm - are they refering to the differance in seg ht ? Reason i ask - is as you all know - when u Look at prism in a PAl- on the 180 ( not fitting or power cross ) - say you have 2.0 base down prism ground - in - and as you know - reguardless of where you move that lens - you will have aprox - the same prism -
***1****
so just wanted some info on the referance of 1mm when dealing in prism in a pal -
what actactly - r u looking for - and how woudl you accomplish this -

** never thougt i would question myself - however - just want to be clear here - ***

**** 2 *****
While i am ?'ing tolerances - say you have a rx - with a .25 cyl - perscribed - and you come up at final with a little less the .25 - would you apply the .12 standards - @ 10 degrees - or the 7 Degree standards for the perscribed .25 - ( that is not really there )) -



thanks everyone - for your understanding and pts -
B

[lensgrinder]

For a PAL in the horizontal meridian you are allowed 1mm each eye in a power above 3.37D, any power less than this you are allowed a total of 0.67^ D.

In the vertical meridian you are allowed a difference of 1mm difference in heights on powers above 3.37D and any power below you are allowed 0.33^ D.

Example: If you have one height that measures 21 and another that measures 20 and the power is 4.00 D O.U. then this is within ANSI standards.

You can view a summary of the tolerances here.
But your question would still fall under the 14 degree tolerance, because you are allowed 14 degrees on cylinder power of 0.00D - 0.25D . If you had an
Rx of -3.00 -0.25 X 90 , but it read a -0.37 in the lensmeter then I would apply the next tolerance, not the 0.00 -0.25 but the >0.25 - 0.50.

[rdcoach5]

Lensgrinder,

Your right the method and standard does allow for 0.67 Prism Diopters to be in one eye and zero in the other, I have never allowed that to pass through my lab, and it has never been very difficult to maintain a higher standard. Good catch, but it was intended to be 0.33 in each lens. If you look at the 1964 ANSI standards the standard was clear in that the tolerence was for each lens, I have attached an image from it.

There is no difference to the patient if the total prism is base out .67 diop in right eye only or in both eyes together.

[HarryChiling]

There is no difference to the patient if the total prism is base out .67 diop in right eye only or in both eyes together.

Sorry for the confusion, it's not the amount in each eye but 0.67. I won't pass a vertical imbalance of 0.67 I still go with 0.50 and your right to soem extent it doesn't matter which eye the prism is in except aestheticaly. I just wanted to point out the fact that the intention was that the prism was split in each eye and along the way this spec was lost. I am not sure if it was doen on purpose or not, but it was done. A little while ago I went on a search to obtain all the standards dateing back to the original '64 standard and I am missing only one, but the standards almost directly mimic the rod our profession has taken whihc has facinated me. Although it is a voluntary standard, it has slowly been relaxed from it's original draft, which has IMO lowered it's effectiveness.

To quote a lab, "ANSI is for amatuers"

Since the standard is voluntary why not keep it tough so that meeting the ANSI standard is an acheivement that only the best labs meet. That could be a badge your lab wears like the ABO is for some opticians. The original standard even had a standard in it for off axis power. Can you imagine that off asxis power, them old cats were really professionals. I was thinking of making a summary of each standard and putting them next to each other like a timelien so that the various changes can be compared to each other.

[Barry Santini]

Try applying this Horizontal prism tolerance to Wrap eyewear...

It really needs to be revisted!

Barry

[HarryChiling]

Try applying this Horizontal prism tolerance to Wrap eyewear...

It really needs to be revisted!

Barry

I remember you quoteing a case where the prescribed prism and the tollerance was off just enough to make a wrap Rx off. I agree that wrap Rx's don't necessarily fit this standard, but if you look at the older standards wrap eyewear wouldn't even be possible due to the higher off axis errors. I want to say it was around the '72 standard that they made a section that allowed for exceptions, but I belive it was mainly inserted for isekonic lens forms, but I do believe that wrap eyewear is becomeing popular enough that the processes in place to produce good wrap eyewear shoudl ahve a stricter tolerance than dress, especially since there are other comprimises to overcome in the eyewear due to changes in form and the prescription plane.

[Barry Santini]

Not a "stricter" tolerance, but rather an understanding that the tolreance is not longer equal, i.e., more BASE IN is OK, but less is NOT!

Barry

[HarryChiling]

Not a "stricter" tolerance, but rather an understanding that the tolreance is not longer equal, i.e., more BASE IN is OK, but less is NOT!

Barry

In a sense stricter but I get what you mean and I remember that was the case with your wrap that they were going by a 0.33D prism tolerance per eye and that your wrap required a 0.25D BI so you could get a progressive without any prism or even slightly BO in it and they would still consider that to be within tolerance.

[Darryl Meister]

Just a few comments:

1. The ANSI power tolerances have remain virtually unchanged since 1972 for spherical lens designs (well, the power tolerances went from 0.12 D to 0.13 D). Otherwise, the only significant changes since the 1979 standard were a change in the cylinder axis tolerances to be more consistent with the power tolerances and the addition of separate tolerances for progressive lenses.

2. Prior to 1972, neither plastics nor tempered glass lenses were in use. Moreover, the power tolerances in the original 1964 standard (0.06 D) actually pushed the envelope of what a traditional surfacing process could theoretically deliver in the first place, even if all other factors were spot on, since the precision of a surfacing process utilizing eighth-diopter tooling is only 0.0625 D. Not to mention the argument against the rationale of having a fabrication tolerance that was 4 times tighter than the just-noticeable-difference of the observer and smallest increment typically used during the actual refraction (0.25 D).

3. Several of the tolerances and test methods in the standard were actually tightened in 1995, to reflect ISO standards. However, the standard was universally rejected by much of the industry, including many laboratory technicians. Consequently, the standard was subsequently revised in 1999 in order to revert back to many of the original 1979 tolerances and test methods. The fact is, as the years go by, fewer and fewer eye care professionals are even familiar with some of these more advanced test methods, such as the application of principal meridian tolerances.

4. I'm glad to hear that there are those out there who expect higher quality. And it is certainly your prerogative to hold your laboratory to tighter standards, if they are capable of doing so on a consistent basis. But keep in mind that these tolerances represent a "consensus" standard agreed upon by numerous organizations. So, while you may arbitrarily decide that you want your lenses to meet 0.03 D tolerances, instead of 0.13 D tolerances, it may be a tough sell to the person actually making your lenses. Of course, in a perfect world, we wouldn't need things like "quality guidelines" and "manufacturing tolerances." Your lab may be able to make you lenses to tighter tolerances, but it will certainly impact their margins and, eventually, your cost.

5. Wrap prism should be treated as prescribed prism, and ANSI tolerances should be applied as usual to the prescribed prism values.

6. ANSI is for amateurs. ANSI is for everybody. Also, ANSI standards are industry performance guidelines. While it would be nice for certain laboratories to differentiate themselves by offering better quality, an ANSI standard isn't really the vehicle with which to communicate this benefit. The time, effort, and cost involved in developing ANSI standards, as well as the consensus nature of these standards, preclude such a limited scope.

[chip anderson]

What's tolerance?

[Barry Santini]

Just a few comments:

...since the precision of a surfacing process utilizing eighth-diopter tooling is only 0.0625 D. Not to mention the argument against the rationale of having a fabrication tolerance that was 4 times tighter than the just-noticeable-difference of the observer and smallest increment typically used during the actual refraction (0.25 D).

5. Wrap prism should be treated as prescribed prism, and ANSI tolerances should be applied as usual to the prescribed prism values.

Darryl,

I respectfully disagree to an extent with the above observations:

1.Any dispenser who has worked with his clients critically knows that a *sweet-spot* centered Rx can usually handle ANSI tolerances, and remain "acceptable". However, when a client's focus or accomodative balance is "at the edge of their sweet spot", then a 0.12D difference is more than just noticaeble. I believe that 0.13D is more than an acceptable tolerance, but perhaps this same standard of precision should now be applied to the refraction as well (assuming we can all agree that a refraction is a necssary part of an eye exam....but that's another story....).

2. Wrap prism should not, IMHO, be treated the same as prescribed prism. The acceptable range of values is not *equivalent*, that is, equal plus-and-minus on each side of the intended value. Since wraps require base-in (and NOT base out), we should begin to recognize, like you've stated, that ANSI standards are arrrived at through a consensus...but that the present consensus was reached with a premise applyin g these standards to*flat* lenses (i.e., phroptor/trial lenses, and modest if nonel frame wrap angles;i.e., < 8 degress).

The bottom line is that, in absence of prescription notation to define, communicate or quantify phorias, we *cannot* have a situation where a requested 0.37D Base In O.U. (prescribed?) wrap prism value is almost wholly negated by an ANSI tolerance of 0.67D)

Wrap eyewear really is different from our training. And let's not get into the whole subject of *validating* the MRP for a finished pair of wrap eyewear. Just what methodalogy (that we all have access to) can we agree upon for tolerancing this parameter? Without consensus here, there is no way to tolerance prism values. For me, parallax is always rearing it ugly head when I try to validate the MRP of wrap eyewear

Add'l thoughts?

Barry

[Darryl Meister]

Wrap prism should not, IMHO, be treated the same as prescribed prism. The acceptable range of values is not *equivalent*, that is, equal plus-and-minus on each side of the intended value. Since wraps require base-in (and NOT base out),

While wrap prism is often base in, the prismatic effect perceived by the wearer is identical to the prismatic effect produced by normal unwanted prism. There is no reason to apply the tolerance differently, unless you want to make the same argument for traditionally prescribed prism. Secondly, wrap prism technically is prescribed prism, because it is a prismatic component you are ordering that differs from the normal optical centration distance. Compensated prescription changes, often used for free-form progressive lenses, which are also based on tilt and wrap values, are also treated as prescribed changes in the ANSI standard.